Icosahedral Particle Reconstruction Software

描述（由申请人提供）：许多球形病毒和一些酶复合物形成具有二十面体对称性的颗粒。电子低温显微镜和单粒子重建的使用使我们能够可视化这种类型的粒子的三维结构，分辨率为7-9 A。该分辨率足以分辨长α螺旋并识别大β折叠。在过去的三年中，我们在这个分辨率范围内测定了六种二十面体粒子的结构，其尺寸从600 - 1250埃不等。它们包括疱疹病毒衣壳、水稻矮缩病毒、P22噬菌体的原衣壳和成熟形式、轮状病毒和胞质型多角体病毒。有了这些经验，我们已经确定了一些可能的改进，在套件的软件特定的二十面体颗粒重建，以提高其未来的维护和升级的便利性，确定的结构的准确性，易于使用的生物学最终用户和计算效率。在这份提案中，我们将有四个具体目标。第一个目标是重写C++中的主要代码，包括二十面体粒子中心和取向细化以及傅立叶贝塞尔重构。它们将针对共享和分布式计算机平台进行优化，并将以模块形式编写，以便于维护和修改。第二个目标是分别改进粒子取向参数细化、对比度传递函数和B因子校正以及傅里叶-贝塞尔重建的算法。我们期望这些改进的代码将导致更准确的二十面体颗粒的结构测定。第三个目标是将改进的代码与Python语言绑定器集成在一起，并提供用户友好的图形界面，并输出相关的中间数据处理结果，使用户能够决定改进是否正在走向正确的结构收敛。第四个目标是通过帮助菜单充分记录这些守则，并通过互联网接入广泛和自由地传播这些守则。在软件发布之前，我们将通过模拟和实验数据测试来确认其可靠性和准确性。实验数据包括我们先前确定的亚纳米分辨率的结构，其中一些组分的晶体结构是已知的。

DESCRIPTION (provided by applicant): Many spherical viruses and some enzyme complexes form particles having icosahedral symmetry. The use of electron cryomicroscopy and single particle reconstruction has allowed us to visualize the 3-dimensional structures of this type of particle to 7-9 A resolution. This resolution is sufficient to resolve long alpha helices and recognize large beta sheets. In the past three years, we have determined 6 structures of icosahedral particles with sizes ranging from 600 -1250 A in this resolution range. They include herpesvirus capsid, rice dwarf virus, procapsid and mature forms of P22 bacteriophage, rotavirus and cytoplasmic polyhedrosis virus. With this experience, we have identified a number of possible improvements in the suite of software specific to the icosahedral particle reconstruction in order to improve the ease of its future maintenance and upgrade, the accuracy of determined structure, the ease of usage by biology end-users and the computational efficiency. In this proposal, we will have four specific goals. The first aim is to rewrite the major codes in C++, which includes the icosahedral particle center and orientation refinement and the Fourier Bessel reconstruction. They will be optimized for both shared and distributed computer platforms and will be written in modular form to make them easy to maintain and modify. The second aim is to improve the algorithms for particle orientation parameter refinement, contrast transfer function and B factor corrections and Fourier Bessel reconstruction respectively. We expect that these improved codes will lead to a more accurate structure determination of icosahedral particles. The third aim is to integrate the improved codes with a Python language binder together with a user-friendly graphical interface and to output relevant intermediate data processing results that would allow the users to decide if the refinement is heading to the correct structure convergence. The fourth aim is to document the codes adequately with a help menu and to disseminate them broadly and freely through the internet access. Before the software release, we will confirm their reliability and accuracy by testing with simulated and experimental data. The experimental data consist of structures that we have previously determined to subnanometer resolution and for which the crystal structures of some of their components are known.